Question

I am trying to write a regex using perl but I need some help.
What I want to do is the following, suppose that I have these texts as examples:

1- [NP some/NN text/NNP here/NNP]

I am interested in the words labelled by /NNP, so I want my regex to search every line until it finds:
[NP then a space then (may or may not find) a word labeled with /NN then one or more words labelled with /NNP (and would contain some special characters).

I want to extract the words labelled with /NNP from each line so the result would be:

1- text here

what I have done so far is to extract the labelled words with /NNP from all the examples

while ($line =~ m/\s(\S*?)\/NNP/gs)
{
        my $word = $1;
        print $word." ";
} 
print "\n";

Any Ideas pleas?

Answer 1

Perhaps:

    #!/usr/bin/env perl
    use strict;
    use warnings;
    while (<DATA>) {
        while ( m{(\[NP.+?\])}g ) {
        my $piece = $1;
        1 while $piece =~ m{(\w+)/NNP}g and printf "%s ",$1;
        print "\n";
        }
    }
    __DATA__
    1- [NP Almst$Ar/NN Ebd/NNP AlmEz/NNP AbrAhym/NNP] [NP Almhnds/NN hAnY/NNP HjAb/NNP]
    2- [NP xAld/NNP ftH/NNP Allh/NNP] [NP ESAm/NNP $rf/NNP] [NP AlqAhrp/NNP]

And then you asked to be able to skip lines with only one labeled word. For that, I might do:

#!/usr/bin/env perl
use strict;
use warnings;
my @line = ();
while (<DATA>) {
    while ( m{(\[NP.+?\])}g ) {
        my $piece = $1;
        while ( $piece =~ m{(\w+)/NNP}g ) {
            push @line, $1;
        }
        print "@line\n", @line = () if @line && @line > 1;
    }
}
__DATA__
1- [NP Almst$Ar/NN Ebd/NNP AlmEz/NNP AbrAhym/NNP] [NP Almhnds/NN hAnY/NNP HjAb/NNP]
2- [NP xAld/NNP ftH/NNP Allh/NNP] [NP ESAm/NNP $rf/NNP] [NP AlqAhrp/NNP]
3- [Nothing of interest here]